aco whitepaper final
DESCRIPTION
TRANSCRIPT
Following the Patient Protection and Afforda-
ble Care Act’s emphasis on Accountable Care Or-
ganizations (ACOs) and the announcement of the
Medicare Shared Savings Program, an increased
interest has emerged among providers and payers
to create ACOs. To date, little has been published
regarding the types and locations of organizations
adopting principles of accountable care.
As part of an ongoing national study, Leavitt
Partners identified ACOs from news releases, me-
dia reports, trade groups, collaborations and inter-
views through the beginning of September 2011.
Also included were entities that either self-
identified as being an ACOs or specifically adopted
the tenets of accountable care including financial
accountability for the health care needs of a popula-
tion, managing the care of that population and bear-
ing that responsibility at an organizational level.
Leavitt Partners then mapped the market of each of
these entities based on the States and Hospital Re-
ferral Regions (HRR) associated with the hospitals
that each entity utilizes.
Of the 164 identified ACOs, the sponsoring
entities included hospital systems, physician groups
and insurers with a market presence in 41 states but
less than half of all HRRs. Of these entities, 99
were primarily sponsored by hospital systems, 38
by physician groups and 27 by insurers.
A clear movement is evolving within the
health care industry towards the accountable care
model of providing health services. Adoption of
this model will vary greatly due to both regional
differences as well as variations among the spon-
soring entities.
Since the 2010 passage of the Patient Protec-
tion and Affordable Care Act (PPACA), industry,
media and national interest has grown in the con-
cept of the Accountable Care Organization (ACO)
1,2. With backing from the White House3 and the
conviction of Centers for Medicare and Medicaid
Services (CMS) leadership that they will lead to
better care, better results and decreased costs4,
Medicare has placed added emphasis on develop-
ing ACOs as part of the Shared Savings Program5
and Pioneer ACO6 demonstration projects. Addi-
tionally, private payers are experimenting with
ACO-centric initiatives in an effort to increase the
value they receive for the prices they pay by lower-
ing the cost of care, improving the outcomes, or,
ideally, both7.
While there are some specific requirements to
participate in Medicare’s demonstration programs,
ACOs can take many different forms within and
apart from Medicare. Since there are likely many
models that will be able to achieve the same goals,
there is little reason to define what an ACO is and,
instead, the emphasis should be on identifying what
an individual ACO does and then study the differ-
ent approaches that can lead to the desired results.
To this end, the loose definition of an ACO sug-
gested by McClellan et al is the most fitting: an
organization that seeks “per capita improvements
in quality and cost” with some degree of accounta-
bility8. To clarify, an ACO must be, to some ex-
tent, financially accountable for the health care
needs of a population, manage the care of that pop-
ulation and bear that responsibility at an organiza-
tional level.
While the “Accountable Care Organization”
name is of recent devise9, the concepts it embraces
are not new to this period: management of and ac-
countability for health care. From the earliest ex-
periments with capitated payments to the most re-
cent pioneer ACO demonstration programs, the
goal of improving outcomes while providers man-
age some degree of risk has been approached in
many different ways. To date, there is no consen-
sus regarding which models are best, and the amor-
phous concept of what an ACO consists of, what it
is expected to do and how it achieves its aims is yet
to be adequately defined, tested or analyzed.
Leavitt Partners Center for ACO Intelligence has
begun to study the organizations that are attempting
to achieve the aims of an ACO without limiting the
approaches the organizations may take and hope to
learn which, and to what degree, approaches are
successful at improving the value of health care.
Without mandatory accreditation10 or some
minimum requirement to become an ACO, Leavitt
Partners has sought to pinpoint ACOs by identify-
ing two types of organizations: those that self-
identify as ACOs and those who have been specifi-
cally identified as adopting the tenets of accounta-
ble care. Leavitt Partners has used news releases,
media reports, trade groups, collaborations, inter-
views and contacts within organizations through
the beginning of September 2011 to identify 164
ACO entities, including those that are actively
bearing risk and coordinating care and those that
are implementing such programs. Initial review
shows large variability between ACO organiza-
tions: some organizations have been bearing risk
and coordinating care for decades, while others
have newly adopted the ACO model and are in an
implementation phase; some ACOs are started by
hospitals and others by physician groups or insur-
ance companies; some are large integrated systems
and others are smaller and primarily variations on
the patient-centered medical home. Starting with
this initial list, Leavitt Partners will continue to
track these and future ACOs over time and evaluate
the effectiveness of different approaches to achiev-
ing the goals of improving care and lowering cost.
This paper addresses the geographic growth of
ACOs in the United States and summarizes the
types of organizations that are implementing the
ACO model. This information is useful as it indi-
cates the regions which should expect initial ACO
growth and describes the types of entities that will
drive the initial creation of ACOs.
1) Dispersion of accountable care organiza-
tions varies significantly by market. There is
extreme variation in the present growth of ac-
countable care organizations with some markets
having multiple ACOs with others having
none. Much ACO growth appears to be a reac-
tion to other organizations in the market: when
one institution forms an ACO, its competitors
often follow suit.
2) Certain regions of the United States are
devoid of accountable care organiza-
tions. While ACO growth is extensive in some
regions, others have no current ACO activi-
ty. Poorer and rural regions in particular have
little ACO growth.
3) Hospitals and hospital systems are the pri-
mary backers of ACOs. Nearly two-thirds of
ACOs identified were started by hospitals or hos-
pital systems. Insurers and Physician Groups,
though, are also adopting tenets of accountable
care and are backing ACOs throughout the coun-
try. The multitude of entities creating ACOs
have led to many different models of providing
care for a patient population.
4) Significant investment in the accountable
care model exists independent of the Medicare
Shared Savings Program. Though the Medi-
care Shared Savings Program final regulations
have been released, implementation is still in its
infancy. Regardless, ACO growth is growing
independent of Medicare as multiple entities
throughout the country are already operating un-
der accountable care payment contracts.
5) The success of different accountable care
models is yet unproven. The overriding goal of
accountable care organizations is to lower costs,
improve care, or both. While there are many
different models of providing accountable care,
which approaches are most successful at realiz-
ing an ACO’s goals is still unclear.
Health care delivery in America is still pri-
marily a cottage industry with few national
health care providers. Most health service pro-
viders are regional and are focused around one
market area, whether because of the simplicity of
dealing with one state law, the difficulties in ex-
panding beyond a relatively small footprint or for
other reasons. Figure 1 depicts the dispersion of
ACOs at the state level. Leavitt Partners classi-
fied state coverage based on the location of hos-
pitals affiliated with the ACO. Where ACOs
cover multiple states, both states were depicted
on the map. When the geographic boundaries
were unclear, as was often the case with large
insurance companies, those ACOs were not in-
cluded on the map; of the 140 ACOs mapped,
127 did not extend beyond one state.
Generally, states with larger populations are
associated with more ACOs, though the trend in
the South, through the plains states and into the
mountain west is toward fewer ACOs. There
are also noticeable outliers such as Montana, the
45th most populous state, which has the same
number of ACOs (three) as Illinois and Georgia,
the 5th and 9th most populous state, respectively.
An indicator of competition among provid-
ers is the number of ACOs in Hospital Referral
Regions (HRRs). Developed by the Dartmouth
Institute for Health Policy, the 306 HRRs are
regional health care markets where patients are
referred for tertiary care11,12. Multiple ACOs in a
single HRR is indicative of markets where health
care providers within the regions may be compet-
ing for the same patients. Figure 2 shows the
number of ACOs by HRR, determined by the
location of hospitals affiliated with the ACO.
When an ACO covers multiple HRRs, all were
included on the map. When an ACO covers a
poorly-defined region or is nearly national in
scope, as is the case with some insurance compa-
ny sponsored ACOs, the ACO was excluded
from this map.
The smaller size of HRRs shows the trend
of entities creating ACOs in narrower regions
than the state map suggests: While only nine
states do not have ACOs, less than half of all
HRRs (144 out of 306) have an ACO. This clus-
tering within HRRs suggests that competing
health systems are simultaneously creating
ACOs. This may arise from providers in a mar-
ket who seek to match or copy what a competitor
is doing or it may be indicative of previously-
integrated systems that are better prepared to be-
come ACOs. Additionally there likely are mar-
ket-specific reasons that have previously affected
the growth of health care entities in different are-
as of the country which differently affect market-
level ACO growth.
Another interesting aspect of this map is the
dearth of ACOs in the Southeast and Appalachi-
an regions which consistently rank as the least
healthy areas of the country13 with a high preva-
lence of obesity, heart disease, diabetes and other
chronic diseases14. Accordingly, it would seem
that these regions stand to benefit the most from
coordinated care15. The reason for the lack of
ACOs in these regions is unclear.
Traditional approaches to coordinated care
have been structured around hospital systems or
payers affiliated with hospital systems. ACOs,
though, can be started by any entity that is able to
cover a large number of lives and bear some
form of risk for that population. Leavitt Partners
defined the sponsoring entity as the organization
that is primarily responsible for the ACO. In
evaluating the sponsoring entity, each entity was
defined as a hospital or health system, an inde-
pendent physician association (IPA) or as an in-
surer. In actuality, some ACOs were started by
organizations that do not clearly fit into one of
these three categories and others were formed as
joint ventures. In seeking to simplify the classifi-
cations, each organization was classified by the
entity that was predominantly responsible for the
ACO’s creation and grouped the ACOs based on
the state where the sponsoring entity is headquar-
tered. Table 1 shows the breakdown of the num-
ber of ACOs formed by each sponsoring entity.
There is a clear trend toward hospital systems
sponsoring ACO development, as they accounted
for more than 60% of all sponsoring entities.
State
Hospital
System
IPA Insurer Total
Alabama 0 0 0 0
Alaska 0 0 0 0
Arizona 2 1 0 3
Arkansas 0 0 1 1
California 8 7 2 17
Colorado 1 0 1 2
Connecticut 1 0 1 2
Delaware 0 0 0 0
D.C. 0 0 0 0
Florida 2 1 0 3
Georgia 1 2 0 3
Hawaii 1 0 0 1
Idaho 0 0 0 0
Illinois 3 1 1 5
Indiana 1 0 1 2
Iowa 1 0 0 1
Kansas 0 0 0 0
Kentucky 2 1 0 3
Louisiana 1 0 0 1
Maine 1 0 1 2
Maryland 2 0 2 4
Massachusetts 5 2 2 9
Michigan 8 3 1 12
Minnesota 2 2 3 7
Mississippi 0 0 0 0
Missouri 3 0 0 3
State
Hospital
System
IPA Insurer Total
Montana 1 0 2 3
Nebraska 1 0 0 1
Nevada 0 0 0 0
New Hampshire 2 0 0 2
New Jersey 5 3 1 9
New Mexico 1 1 1 3
New York 4 3 1 8
North Carolina 2 2 1 5
North Dakota 0 0 0 0
Ohio 7 0 0 7
Oklahoma 1 0 0 1
Oregon 2 1 0 3
Pennsylvania 4 0 2 6
Rhode Island 0 0 0 0
South Carolina 1 0 0 1
South Dakota 0 0 0 0
Tennessee 2 1 2 5
Texas 8 4 0 12
Utah 1 0 0 1
Vermont 0 0 0 0
Virginia 1 0 0 1
Washington 3 3 0 6
West Virginia 0 0 0 0
Wisconsin 7 0 1 8
Wyoming 1 0 0 1
Total 99 38 27 164
With the Medicare Shared Savings Program
still to be implemented, the substantial growth of
Accountable Care Organizations indicates a trend
within the health care industry towards the account-
able care model, partially independent of govern-
ment incentives. With significant regional varia-
tion, it is unclear, though, what is driving market-
level ACO growth. In some large markets, such as
Boston, ACOs are proliferating, while in other large
markets, such as Washington DC, they are not.
Market specific clustering is a prevalent feature—if
there is one ACO, it is more likely that another is
nearby. Further tracking of ACO growth and dis-
persion will provide a more sound conclusion as to
whether ACO adoption is primarily a response to
competitors, indicated by future ACO growth re-
maining concentrated around existing ACOs, or
indicative of the success and effectiveness of the
model, thereby dispersing throughout all markets.
As a consensus regarding the definition of an
ACO continues to develops, evidence exists that the
basic tenets of accountable care have existed in
many organizations for years, and only the title of
ACO is new. Preliminary review of the organiza-
tions we have identified indicates a trend toward
proclaiming oneself as an ACO with only modest
changes to the care process, rather than radically
redesigning the organization to become something
fundamentally different in the future. It appears,
for now, that defining oneself as an ACO represents
an acceptance of the direction the industry has been
headed rather than an adoption of a truly new form
of care delivery.
The range of entities that have sponsored
ACOs, from small IPAs to national insurance com-
panies indicates the wide range of business models
that will ultimately provide accountable care. Un-
der the Shared Savings Program, entities must be
care providers to qualify16, but non-provider insur-
ance companies are a major backer of ACO growth,
indicating a much broader definition of what type
of entity can provide accountable care. Important
insights will be drawn by observing which models
succeed in reaching the overriding goal of increas-
ing value through improving quality, lowering costs
or both.
With neither a set definition nor a national method
for identifying ACOs, it is difficult to precisely
identify and study such organizations. It is possible
that some of the organizations which should be
considered ACOs are missing from our study and
some, such as organizations that self-identify as
ACOs but will never ultimately adopt any type of
care coordination or bear any risk for a population,
may not belong. Accurate representation of all
ACOs will happen with further analysis of the cur-
rent organizations on our list and future identifica-
tion of other ACO entities.
There are also limitations with mapping where the
ACO is located. The geographic area covered by
an ACO is not always clear, leading to possibly
inaccurate depictions of the geographic dispersion
of ACOs. For example, some sponsoring organiza-
tions have a population they presently serve, but
the ACO they have announced may only exist in
part of the region that the sponsoring organization,
as a whole, covers. Additionally, some ACOs are
organized by regional or national entities that may
cover ill-defined patient populations in many states,
making completely accurate determination of the
geographic region that the ACO covers unknowa-
ble.
1 Goldsmith, Jeff. “Accountable Care Organizations: The Case For Flexible Partnerships Between Health
Plans And Providers,” Health Affairs 30, no. 1 (January 1, 2011): 32 -40; Steven M. Lieberman and John
M. Bertko, “Building Regulatory And Operational Flexibility Into Accountable Care Organizations And
‘Shared Savings’,” Health Affairs 30, no. 1 (January 1, 2011): 23 -31.
2 “National Accountable Care Organization Congress: Overview”, n.d., http://www.acocongress.com/
overview.html.
3 HealthCare.gov. “Accountable Care Organizations: Improving Care Coordination for People with Medi-
care”, March 31, 2011, http://www.healthcare.gov/news/factsheets/2011/03/accountablecare
03312011a.html.
4 Berwick, Donald M. “Launching Accountable Care Organizations — The Proposed Rule for the Medicare
Shared Savings Program,” New England Journal of Medicine 364 (April 21, 2011): e32.
5 Centers for Medicare & Medicaid Services. “Overview of the Shared Savings Program”, May 17, 2011,
https://www.cms.gov/sharedsavingsprogram/.
6 Center for Medicare & Medicaid Innovation. “Pioneer ACO Model”, n.d., http://innovations.cms.gov/areas
-of-focus/ seamless-and-coordinated-care-models/pioneer-aco/.
7 Delbanco, Suzanne F., et al. Promising Payment Reform: Risk-Sharing with Accountable Care Organiza-
tions (The Common wealth Fund, July 2011), http://www.commonwealthfund.org/~/media/Files/
Publications/Fund%20Report/ 2011/Jul/1530Delbancopromisingpaymentreformrisksharing%202.pdf.
8 McClellan, Mark, et al. “A National Strategy To Put Accountable Care Into Practice,” Health Affairs 29,
no. 5 (May 1, 2010): 982-990.
9 Medicare Payment Advisory Commission. Public Meeting, November 8, 2006, http://www.medpac.gov/
transcripts/1108_1109_medpac.final.pdf.
10 The National Committee for Quality Assurance (NCQA) is implementing a voluntary accreditation process;
see “Accountable Care Organization Accreditation”, n.d., http://www.ncqa.org/tabid/1312/Default.aspx.
11 For more information on HRRs, please see Dartmouth Institute for Health Policy and Clinical Practice.
“Dartmouth Atlas of Health Care”, n.d., http://www.dartmouthatlas.org/.
12 Geographic boundary files for HRRs were obtained from Dartmouth Atlas of Health Care. “Downloads”,
n.d., http://www.dartmouthatlas.org/tools/downloads.aspx#boundaries.
13 United Health Foundation. “America’s Health Rankings”, n.d., http://www.americashealthrankings.org/.
14 Center for Disease Control. “Behavioral Risk Factor Surveillance System”, August 19, 2011, http://
www.cdc.gov/brfss/.
15 Peikes, Deborah, et al. “Effects of Care Coordination on Hospitalization, Quality of Care, and Health Care
Expenditures Among Medicare Beneficiaries,” JAMA: The Journal of the American Medical Association
301, no. 6 (February 11, 2009): 603 -618.
16 Centers for Medicare & Medicaid Services. Proposed Rule: Medicare Shared Savings Program, 42 CFR
Part 425, 2011, http://www.ftc.gov/opp/aco/cms-proposedrule.PDF.